models.py

import os
from pydantic import BaseModel, Field

from dotenv import load_dotenv
from prompts import *

from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate

load_dotenv()

class RouterResponse_2(BaseModel):
    route :list[str]= Field(description=("A list of keys relevant to the user's query"))
class ExtractorResponse_2(BaseModel):
    information: str=Field(description=("Condensed information based on the context provided"))

llm = ChatOpenAI(model="openai/gpt-4o-mini",temperature=0.7,base_url="https://openrouter.ai/api/v1",
api_key=os.getenv("OPEN_ROUTER_API_KEY"))

router_prompt_1 = ChatPromptTemplate.from_messages([
            ("system", "You are a routing assistant."),
            ("user", router_instruction_prompt_1.format(query="{query}", previous_messages="{previous_messages}",
                                               format_instructions="{format_instructions}"))])
router_chain_1= router_prompt_1 | llm
summary_prompt_1 = ChatPromptTemplate.from_messages([
            ("system", "You are a Summarising assistant."),
            ("user", summary_prompt_instructions_1.format(query="{query}", previous_messages="{previous_messages}",
                                                        data="{data}",format_instructions="{format_instructions}"))])
summary_chain_1 = summary_prompt_1 | llm

router = router_instruction_prompt_2 | llm.with_structured_output(RouterResponse_2)
extractor = extract_prompt_instructions_2 | llm.with_structured_output(ExtractorResponse_2)